Disk reporting low IOPS and MBPS

Asset ID:	1-72-2268148.1
Update Date:	2017-05-21
Keywords:

Solution Type Problem Resolution Sure

Solution 2268148.1 : Disk reporting low IOPS and MBPS - ODA

Applies to:

Oracle Database Appliance X3-2 - Version All Versions to All Versions [Release All Releases]
Information in this document applies to any platform.

Symptoms

Newly deployed 12.1.2.9.0 ODA X3-2,

running 'oakcli validate -c DiskCalibration' shows one SAS disk having just over half the MB/s as the rest:

RESULT: Random read throughput of HDD_E0_S15_376391396 is 234 IOPS 147 MBPS
RESULT: Random read throughput of HDD_E0_S17_376391424 is 238 IOPS 153 MBPS
RESULT: Random read throughput of HDD_E0_S07_376390852 is 235 IOPS 148 MBPS
RESULT: Random read throughput of HDD_E0_S04_376390396 is 227 IOPS 147 MBPS
RESULT: Random read throughput of HDD_E0_S01_376391436 is 238 IOPS 151 MBPS
RESULT: Random read throughput of HDD_E0_S05_376762192 is 237 IOPS 147 MBPS
RESULT: Random read throughput of HDD_E0_S10_376762548 is 204 IOPS 130 MBPS
RESULT: Random read throughput of HDD_E0_S09_376718004 is 230 IOPS 144 MBPS
RESULT: Random read throughput of HDD_E0_S12_376762224 is 234 IOPS 152 MBPS
RESULT: Random read throughput of HDD_E0_S18_376762524 is 237 IOPS 148 MBPS
RESULT: Random read throughput of HDD_E0_S14_376762520 is 236 IOPS 150 MBPS
RESULT: Random read throughput of HDD_E0_S00_376391324 is 232 IOPS 145 MBPS
RESULT: Random read throughput of HDD_E0_S13_376515908 is 234 IOPS 149 MBPS
RESULT: Random read throughput of HDD_E0_S06_376762276 is 238 IOPS 149 MBPS
RESULT: Random read throughput of HDD_E0_S03_376516984 is 235 IOPS 148 MBPS
RESULT: Random read throughput of HDD_E0_S19_376391432 is 231 IOPS 151 MBPS
RESULT: Random read throughput of HDD_E0_S16_376391480 is 224 IOPS 82 MBPS <RESULT: Random read throughput of HDD_E0_S08_376390936 is 229 IOPS 146 MBPS
RESULT: Random read throughput of HDD_E0_S11_376390800 is 232 IOPS 147 MBPS
RESULT: Random read throughput of HDD_E0_S02_376391428 is 233 IOPS 149 MBPS

The following command's output indicate the disk looks fine.

# oakclistordiag e0_pd_16 --> No issues

# smartctl --all /dev/sdao --> No issues

There is no direct errors for the disk was seen, reboot node does not help.

Changes

Newly deployed 12.1.2.9.0 ODA X3-2

Cause

There is a good chance that the disk is failing if the disk has 1/2 iops that is a big potential problem as the entire diskgroup will default to the slowest disk even if the disk has not yet completely failed

Solution

1. Replace the disk

2. Run calibration

After changing the disk, the IOPS and MBPS are back to normal similar to the rest of the disks

Attachments

This solution has no attachment